Hugging Face's logo

btemirov
/
distill-whisper-jargon

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
whisper
Generated from Trainer
Model card Files Metrics Community
btemirov commited on
Commit
ade5bfa
·
verified ·
1 Parent(s): 7e1e8df

Model save

Browse files
Files changed (3) hide show
  1. README.md +68 -0
  2. generation_config.json +195 -0
  3. model.safetensors +1 -1
README.md ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ base_model: distil-whisper/distil-small.en
4
+ tags:
5
+ - generated_from_trainer
6
+ metrics:
7
+ - wer
8
+ model-index:
9
+ - name: distill-whisper-jargon
10
+ results: []
11
+ ---
12
+
13
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
14
+ should probably proofread and complete it, then remove this comment. -->
15
+
16
+ # distill-whisper-jargon
17
+
18
+ This model is a fine-tuned version of [distil-whisper/distil-small.en](https://huggingface.co/distil-whisper/distil-small.en) on the None dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 4.4314
21
+ - Wer: 78.4173
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 1e-05
41
+ - train_batch_size: 4
42
+ - eval_batch_size: 8
43
+ - seed: 42
44
+ - gradient_accumulation_steps: 4
45
+ - total_train_batch_size: 16
46
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
47
+ - lr_scheduler_type: linear
48
+ - lr_scheduler_warmup_steps: 200
49
+ - training_steps: 500
50
+ - mixed_precision_training: Native AMP
51
+
52
+ ### Training results
53
+
54
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
55
+ |:-------------:|:------:|:----:|:---------------:|:-------:|
56
+ | 3.8199 | 22.22 | 100 | 3.5227 | 79.3525 |
57
+ | 2.3504 | 44.44 | 200 | 3.7073 | 77.3022 |
58
+ | 1.4612 | 66.67 | 300 | 4.1042 | 78.6691 |
59
+ | 0.9713 | 88.89 | 400 | 4.3164 | 77.7698 |
60
+ | 0.7453 | 111.11 | 500 | 4.4314 | 78.4173 |
61
+
62
+
63
+ ### Framework versions
64
+
65
+ - Transformers 4.36.2
66
+ - Pytorch 2.1.0+cu121
67
+ - Datasets 2.16.1
68
+ - Tokenizers 0.15.0
generation_config.json ADDED
@@ -0,0 +1,195 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alignment_heads": [
3
+ [
4
+ 6,
5
+ 6
6
+ ],
7
+ [
8
+ 7,
9
+ 0
10
+ ],
11
+ [
12
+ 7,
13
+ 3
14
+ ],
15
+ [
16
+ 7,
17
+ 8
18
+ ],
19
+ [
20
+ 8,
21
+ 2
22
+ ],
23
+ [
24
+ 8,
25
+ 5
26
+ ],
27
+ [
28
+ 8,
29
+ 7
30
+ ],
31
+ [
32
+ 9,
33
+ 0
34
+ ],
35
+ [
36
+ 9,
37
+ 4
38
+ ],
39
+ [
40
+ 9,
41
+ 8
42
+ ],
43
+ [
44
+ 9,
45
+ 10
46
+ ],
47
+ [
48
+ 10,
49
+ 0
50
+ ],
51
+ [
52
+ 10,
53
+ 1
54
+ ],
55
+ [
56
+ 10,
57
+ 2
58
+ ],
59
+ [
60
+ 10,
61
+ 3
62
+ ],
63
+ [
64
+ 10,
65
+ 6
66
+ ],
67
+ [
68
+ 10,
69
+ 11
70
+ ],
71
+ [
72
+ 11,
73
+ 2
74
+ ],
75
+ [
76
+ 11,
77
+ 4
78
+ ]
79
+ ],
80
+ "begin_suppress_tokens": [
81
+ 220,
82
+ 50256
83
+ ],
84
+ "bos_token_id": 50257,
85
+ "decoder_start_token_id": 50257,
86
+ "eos_token_id": 50256,
87
+ "forced_decoder_ids": [
88
+ [
89
+ 1,
90
+ 50362
91
+ ]
92
+ ],
93
+ "is_multilingual": false,
94
+ "language": "<|en|>",
95
+ "max_initial_timestamp_index": 1,
96
+ "max_length": 448,
97
+ "no_timestamps_token_id": 50362,
98
+ "pad_token_id": 50256,
99
+ "return_timestamps": false,
100
+ "suppress_tokens": [
101
+ 1,
102
+ 2,
103
+ 7,
104
+ 8,
105
+ 9,
106
+ 10,
107
+ 14,
108
+ 25,
109
+ 26,
110
+ 27,
111
+ 28,
112
+ 29,
113
+ 31,
114
+ 58,
115
+ 59,
116
+ 60,
117
+ 61,
118
+ 62,
119
+ 63,
120
+ 90,
121
+ 91,
122
+ 92,
123
+ 93,
124
+ 357,
125
+ 366,
126
+ 438,
127
+ 532,
128
+ 685,
129
+ 705,
130
+ 796,
131
+ 930,
132
+ 1058,
133
+ 1220,
134
+ 1267,
135
+ 1279,
136
+ 1303,
137
+ 1343,
138
+ 1377,
139
+ 1391,
140
+ 1635,
141
+ 1782,
142
+ 1875,
143
+ 2162,
144
+ 2361,
145
+ 2488,
146
+ 3467,
147
+ 4008,
148
+ 4211,
149
+ 4600,
150
+ 4808,
151
+ 5299,
152
+ 5855,
153
+ 6329,
154
+ 7203,
155
+ 9609,
156
+ 9959,
157
+ 10563,
158
+ 10786,
159
+ 11420,
160
+ 11709,
161
+ 11907,
162
+ 13163,
163
+ 13697,
164
+ 13700,
165
+ 14808,
166
+ 15306,
167
+ 16410,
168
+ 16791,
169
+ 17992,
170
+ 19203,
171
+ 19510,
172
+ 20724,
173
+ 22305,
174
+ 22935,
175
+ 27007,
176
+ 30109,
177
+ 30420,
178
+ 33409,
179
+ 34949,
180
+ 40283,
181
+ 40493,
182
+ 40549,
183
+ 47282,
184
+ 49146,
185
+ 50257,
186
+ 50357,
187
+ 50358,
188
+ 50359,
189
+ 50360,
190
+ 50361
191
+ ],
192
+ "task": "transcribe",
193
+ "transformers_version": "4.36.2",
194
+ "use_scan": false
195
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:50d7c152bfc1d329eec2fe9eff5c2bdff14b8f38aed5c89284fc756de6fba7a1
3
  size 664561848
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff64e397fade39f806fa093514b0dd643df8754a94d7bd178a623e8790f2df66
3
  size 664561848